Usability testing: Bernard and Belinda’s route to information – Davina Borman & Frankie Wilson, Brunel University Library

Frankie and Davina have been giving their library’s website an overhaul and they have made great progress improving both the content and the compliance with accessibility guidelines. The content of this presentation was a great deal more practically focussed than some of the more theoretical ones that I attended.

They took us through the whole process but the most interesting aspect of it was that referred to in the title – their approach to usability testing. They started with creating some personas based on the different user groups and then asked librarians to ‘get into character’ and perform a few tasks on the new website. Although it was very time consuming, it turned up several aspects of the website that could be improved. Next, they trialled the new site using volunteer undergrads from the University. Although this group identified a few more issues, none of them contradicted those identified as part of the persona process. I’d say that’s a vote in favour of the persona exercise.

I can see the application of this persona concept really coming into its own when testing a website for which the audience is so large that it would be difficult to engage the end user. In this particular project, the audience is a finite and known group: the students. On top of that, it is a group for which they have the means of contacting every individual member. With such a closed group, I would have thought it would make more sense to engage a representative selection of the student body in the process right at the beginning to ensure that the first pass was as close to what students wanted as possible. I would then have used that same group and second representational one to test the site.

Frankie and Davina are the first to admit that they didn’t really know what they were doing and that it was a lot of enthusiasm, reading and a steep learning curve for them, but I think that their approach was a little naïve. Having said that, the methodology that they used appears sound and the results of the student testing did, to a large extent, prove the effectiveness of the persona approach. As the audience for our website is so large both in number and geographical spread and is largely unknown, I will be suggesting to the project team that we could use this persona methodology when we review our website in the next year.

Surveying subscribers

We publish a monthly e-newsletter for which I am responsible (content, creation, distribution, data management, subscriber satisfaction / evaluation). I have done quite a bit to improve on each of the first four aspects and now need to turn my attention to the last. Apparantly, we conducted a survey almost two years ago now so I think that another one is about due. On top of that, I’m interested to know what subscribers think of the changes or whether they have even noticed them. Some of the changes they are not likely to have noticed (such as the changes to the template which ensure that every issue conforms to XHTML, WCAG and CSS guidelines - all links open in new window), some they won’t know about (such as the new software we use to manage subscriber data and e-newsletter distribution) and others they almost certainly will have noticed (such as the new look our e-newsletter has received).

I haven't done a whole lot of thinking on this one yet but I can immediately see three ways forward:

1. The first is to put together a few questions in an email to which we ask subscribers to reply. This is probably the cheapest option but:
   1. it’s probably the least professional and slick looking to the subscribers
   2. it’s more onerous on the subscribers
   3. and certainly creates a fair old bit of work for me in compiling the responses for analysis.
2. The second is to outsource the entire thing to a company specialising in this sort of work. This is likely to yield the most professional-looking result for the subscribers and would be the least onerous for me in terms of data compilation and analysis but is almost certainly going to be the most expensive option.
3. The third option is to use something like surveymonkey.com (opens in new window) which would enable us to create our own online surveys and does the data complilation and some basic analysis for you. It’s the least flexible of the options (you are working within the confines of an OTS product) but it probably offers the best combination of professional appearance, simplicity for subscribers, simplicity for me, and cost.

I have had a go at putting something together use surveymonkey.com (there is a free version available but I think that we would want to use one of the paid subscription versions as the benefits of doing so are in our favour – e.g. limitations on the number of forms submitted are raised or removed depending on which package you opt for) and am happy enough with it.

Has anyone out there done a survey like this before using a tool like surveymonkey.com? What software did you use? Any ‘top tips’ I might use?